Journal of Geotechnical and Geoenvironmental Engineering

The thermo–hydro–mechanical (THM) behaviors of frozen soils are often modeled based on the thermodynamic fluxes of moisture and heat. However, existing models disregard the pore-scale granular interaction between the soil grain and ice crystal for saturated partially frozen soils. The pore-scale mechanism of pore-filling and load-bearing for the ice phase in the loaded soil skeleton has not been explored. An alternative constitutive model is therefore proposed by considering the microscopic temperature-dependent distribution of the ice phase for pore-filling and load-bearing in the soil interpore. This reflects the influence of the ice phase on the soil stress states and the associated THM behaviors as interpreted based on the critical state framework. The model was validated by published experimental results and considerably captured undrained shearing behaviors of frozen soils at various temperatures and confining pressures. A numerical parametric study was conducted to investigate the dependency of the phase relationship, stress state, undrained shear strength, and soil stiffness on temperature. The modeling suggests that the ice crystals filling in the pore are partially load-bearing to affect the soil stiffness and partially unloaded to alter the stress state. It shows that freezing turns the soil into a heavily consolidated state by increasing the specific volume and decreasing the effective granular void ratio. The undrained shear strength of frozen soils increases with a decrease in temperature because the dilatancy is enhanced due to the ice invasion in the interpore. It also demonstrates that soil stiffness is influenced by not only the stress state but the freezing history. This study highlights the temperature-dependency of mechanical behaviors and the validity of using the concept of stress states to interpret the pore-scale mechanistic soil–water–ice interactions for frozen soils.

The effect of particle shape on the suffusion of gap-graded soils is an essential although poorly understood subject in geotechnical engineering that requires further investigation. This work presents a macroscale and microscale numerical investigation into the effect of particle shape on the suffusion of gap-graded granular materials. Rounded, elliptical, and convex particles with the same volume-equivalent diameter and varying shape coefficients were generated and used to produce samples. Next, a series of resolved coupled computational fluid dynamics (CFD) and discrete-element method (DEM) simulations were performed to provide evidence of the effect of particle shape on the suffusion susceptibility of gap-graded soils. The evolution of particle orientation, moment, and drag force coefficient were analyzed to determine the mechanisms by which particle shape exerts influence. The fine angular particles under seepage flow were found to adjust their orientation, reducing the projected area of the particle perpendicular to the fluid flow direction. Fine particles in high-flow-velocity regions had a smaller projected area and drag force coefficient. The continuous rotation of the irregularly shaped particles during suffusion implies that their migration should counteract the moments exerted by the surrounding particles. In the sample containing various irregularly shaped particles, the initial position of the most irregularly shaped particle was closer to the outlet, implying that irregularly shaped particles are less susceptible to suffusion.

This article presents experimental data on the earth pressures that develop between mechanically stabilized earth (MSE) wall panels that are tied together in an unyielding condition. Prior to design and construction, an extensive literature review was conducted to ensure the MSE wall adhered to the local standard specifications for road and bridge construction, current design codes, and the construction sequencing specified by the wall panel manufacturer. The Strong Wall at the University of Florida’s Structures and Materials Laboratory was utilized to construct a model full-scale MSE wall using inextensible reinforcement, with 140 locations continuously monitored. The MSE wall was divided into two different relative compaction zones, 95% and 103% of the reinforced soil’s optimum dry density, to investigate a range of friction angles that can develop during conventional construction. To achieve the necessary aspect ratio of 0.3 specified by practitioners for use in real-world applications, a reaction frame was incrementally loaded on top of the reinforced backfill to simulate additional wall height and overburden stress. Earth pressure coefficients were found for each compacted soil lift and incremental load from the reaction frame for both states of soil density. A force equilibrium analysis was conducted to ensure all forces within the experimental setup were accounted for to validate the results. From the derived earth pressure coefficients, it was observed that earth pressures moved from a passive condition to an active or at-rest condition as the soil-height increased above each reinforcement level. It was concluded that increased lateral stress develops from the compaction effort in an unyielding condition, which is not accounted for using conventional MSE design methods. Consequently, an equation was developed that incorporates a variable friction angle ($\phi$) based on the compaction effort for an unyielding condition that closely followed the trends of the measured results.

Chiaia station is one of the art stations of Line 6 of the Naples underground network; it was constructed in a $50\text{-}\mathrm{m}\text{-}\mathrm{deep}$ excavation, a few meters from historic buildings and 4.5 m from the main façade of a sixteenth-century Basilica. The excavation, carried out partially in loose to medium dense sands overlying the soft rock formation of Neapolitan Yellow Tuff (NYT), was supported by a retaining wall made of contiguous bored piles braced with internal struts and prestressed ground anchors. The excavation sequence was quite complex due to archeological findings and to the presence of anthropic cavities used over the centuries to quarry NYT blocks. One of the key goals of the design was to limit movements around the shaft to prevent damage in the historical buildings. Long-term monitoring data obtained during nearly 9 years confirm the success of the overall construction process. A rather complex three-dimensional (3D) finite-element model with constitutive relationships for both the upper sandy layers and the soft rock is presented in the paper; this model was adopted to back-analyze the data from the monitoring and explore the influence of some of the key features of the case study. The role of the building bending and shear stiffness, of the soft rock stiffness, and of further apparently minor issues—such as the seepage and the ground anchors’ prestress—were investigated and discussed with the support of the model calculations. Observed settlements at the end of the excavation were in the range 10–15 mm, and in the long term they increased by 20%–50% to as much as 20 mm. The deflection ratios were very small, in the range $0.05‰–0.15‰$, and no visible damages to the buildings were recorded. These values were reproduced by the finite-element model only after the introduction of the relevant building stiffness.

Numerous recent failures of tailings storage facilities (TSFs) (e.g., the 2019 Brumadinho failure in Brazil, the 2018 Cadia failure in Australia) have occurred in the last decade, providing important case histories and lessons to improve the design of TSF facilities. This study is focused on the Cadia TSF failure; specifically, we reexamine and expand the triggering analyses conducted in the forensic study after the failure. In addition, this study also presents the first comprehensive posttriggering and runout assessment of the Cadia failure. The triggering mechanisms were evaluated through Lagrangian-based analyses with the finite-difference method (FDM), which is well suited for small to moderate deformations, whereas the posttriggering mechanisms and runout were evaluated using the material point method (MPM), which is well suited for the evaluation of large deformations. The FDM-based analyses show that the interaction between a relatively weak layer in the foundation [Forest Reef Volcanic Unit A (FRV-A)], the deposited tailings, and the construction of a buttress system is critical in explaining the failure triggering. The MPM-based assessments are useful in investigating posttriggering mechanisms, suggesting a retrogressive failure and stress redistribution occurring within the mine tailings. In addition, the MPM-based assessments also explore the influence of key factors on the posttriggering mechanisms and runout. These factors include the tailings liquefaction extent and strength, the failure propagation rate, the strength of the foundation materials, and the buttress construction. The analyses suggest that the foundation strength, mine tailings liquefaction extent, and buttressing affect the runout assessments significantly. In this context, MPM proves to be especially useful in investigating postfailure conditions.

This paper presents a study on the hysteretic damping of cyclic axially loaded piles in sand that would be of particular interest to the design and analysis of offshore wind turbine (OWT) jacket structures. Foundation damping can help to reduce fatigue loads thereby leading to leaner designs and/or extended design life. Damping has been studied in laterally loaded OWT monopiles, but there is little information and guidance for estimating damping for axially loaded piles in an OWT jacket structure. The objective of this study was to (1) document hysteretic damping in axially loaded pipe piles in sands and (2) develop a simple and practical numerical approach for predicting the hysteretic damping (and stiffness) properties for dynamic numerical analysis of the OWT structure. First, existing low-frequency pile load test data are compiled from the literature and analyzed to evaluate the hysteretic damping. Second, a simple quasi-static equivalent linear modeling approach is described that can be used to predict hysteretic damping at the pile head using a conventional static t-z analysis. The axial stiffness and damping ratios predicted for the test piles at the pile head compared very well to the observed pile data when the maximum interface friction was accurately predicted.

Pile foundations supporting high-rise buildings are generally subject to cyclic loading because of dynamic loading. The corrosion of steel materials in pile foundations is another major concern, especially for piles in a marine environment. In this study, a series of cyclic and static loading tests on model piles made of fiber-reinforced polymer (FRP) and seawater–sea sand concrete (SSC) and ended in a rock socket were reported. Three structural configurations (FRP tube–confined, FRP rebar cage–reinforced, and centered FRP rebar–reinforced) were adopted for the model piles. Strain along the depth of the piles was measured using fiber Bragg grating (FBG) optic sensors and an advanced distributed optical sensing technique known as optical frequency domain reflectometry (OFDR). Strain distribution, axial cyclic stiffness, and shaft friction mobilization of the piles under static and different modes of axial cyclic loading were analyzed and explored in detail. The test results indicated that the FRP tube–confined model pile showed higher confinement and cyclic capacity and lower stiffness degradation, leading to relatively more stable behavior. A high level of cyclic loading can cause microcracks to form and grow within the pile material, thereby decreasing pile stiffness. The strain profile of all the piles along the depth appeared to follow a similar trend, and fluctuations at certain points led to failure. Cyclic stiffness showed gains initially when cyclic load conditions were below a certain threshold level but degraded when loading was increased beyond it. Moreover, shaft resistance gradually increased with cycles, causing higher mobilization in the upper portion of the socket. The experimental results have provided the first systematic study on the performance of the FRP-SSC composite model piles ended in rock sockets under axial cyclic and static loadings. This will contribute to development of a potential predictive method for pile settlement and capacity for the better design of rock-socketed piles.

Existing simplified procedures for evaluating soil liquefaction potential or for estimating excess pore pressures during earthquakes are typically based on undrained cyclic tests performed on saturated soil samples under controlled loading and boundary conditions. Under such conditions, the effect of excess pore pressure ($u_e$) dissipation and redistribution to neighboring soil layers cannot be accounted for. Existing simplified procedures treat liquefiable layers as isolated soil layers without any boundary conditions even if dense and loose layers are very thin, permeable, and adjacent to each other. However, redistribution is likely to increase and decrease $u_e$ in the neighboring dense and loose layers respectively. Until now, no procedure short of fully coupled numerical analysis is available to estimate the importance of redistribution. This paper presents an approximate analytical procedure for assessing the effects of $u_e$ redistribution in (1) soil layers that would have liquefied if they were undrained, and (2) soil layers that would have not liquefied even if undrained. It is found that a layer that is initially assumed liquefied under undrained conditions might not even liquefy accounting for the $u_e$ redistribution to neighboring layers. On the other hand, a layer initially assumed to not liquefy can develop significant $u_e$ and can even liquefy due to pore pressure migration from the neighboring layers. Thus, accounting for redistributed $u_e$ is important for liquefaction consequence assessment quantification, particularly in systems that span the depth of these effects like deep foundations. Migration of u toward the tip of a pile can reduce its capacity, even if the tip is embedded in a dense sand layer. On the other hand, if redistribution can result in the reduction of $u_e$ in initially assumed liquefied layers, risks associated with liquefaction might be avoided. A criterion is also developed to evaluate the thicknesses of a layer below which redistribution could prevent liquefaction even if the layer is deemed liquefied according to the existing liquefaction-triggering procedures. Finally, the proposed procedure is illustrated by application to selected shaking events of centrifuge tests involving liquefaction of layered soil profiles. The predictions from the procedure matched the centrifuge test results reasonably.

The primary cause of embankment dam failure is the overtopping of the crest, making dam safety a crucial consideration due to the potential for catastrophic consequences. Dam safety regulations and guidelines are increasingly demanding the enhancement of rockfill dams’ resistance to overtopping and leakages. Assessing the breaching of embankments is essential for conducting risk evaluations and hazard studies. Typically, this assessment is carried out using parametric breach models, which are statistically derived equations based on historical dam failure cases. However, these models often overlook important parameters such as material properties, leading to uncertainties in the results. This research focuses on quantifying the accuracy of breach parameters (breach width, failure time, peak outflow) predicted by various parametric breach models found in the literature. Three prototype embankment dams (homogeneous earthfill dam, homogeneous rockfill dam, and zoned rockfill dam), along with five historical failure cases, are considered to evaluate the predictive capability of the parametric breach models. The study compares the measured and estimated breaching parameter values, aiming to identify the most suitable parametric breach model based on the dam design. The results indicate that the performance of the different equations in estimating breach parameters significantly depends on the characteristics of the dam and reservoir. Additionally, a simplified physical model (DLBreach) is employed to estimate breach parameters for the three prototype dams, and a comparison with results from parametric breach models is presented. In conclusion, accurately predicting breach parameters is vital for improving dam safety. This research highlights the influence of dam and reservoir characteristics on the performance of different parametric breach models. By identifying the most suitable breach models for specific dam designs, this study contributes to the understanding and management of embankment dam failures.

Carbon footprint considerations have become one of the significant concerns in the construction and rehabilitation of public infrastructure. As such, sustainability metrics using carbon footprint are of particular significance to stakeholders of lifeline infrastructure systems, including bridges, under extreme events. This study aimed to assess the sustainability and resilience of a two-span reinforced concrete bridge under liquefaction and associated ground deformations. For that purpose, a comprehensive three-dimensional (3D) nonlinear finite-element (FE) framework combined with the formulation of performance-based earthquake engineering (PBEE) and economic input-output life-cycle assessment approach was developed. In terms of the estimated post-earthquake repair cost, repair time, and carbon footprint, seismic resilience and robustness of the bridge subjected to liquefaction-induced lateral spreading were explored. Within this holistic framework, the variations in PBEE results for bridges founded on different ground conditions were examined. Furthermore, this study delved into the effects of hydraulic conductivity on a bridge’s sustainability and resilience. It is shown that an increase in hydraulic conductivity can noticeably reduce the post-earthquake repair cost, repair time, and carbon footprint, thus improving the robustness and resilience of the bridge. Overall, the derived insights reveal the need to incorporate sustainability and resilience analysis techniques when addressing seismic hazards and associated liquefaction-induced deformations.

This study focuses on the characterization of low-velocity (lower than $100\mathrm{km}/\mathrm{h}$) rockfall impact loads transferring to a thick steel-reinforced concrete (SRC) slab through a protective sand layer. A full-scale experimental test campaign was performed, where each test consisted of releasing a concrete block which, after a vertical free fall, impacts a sand protective layer placed over a SRC slab, in order to represent an isolated rockfall impact to which an actual SRC structure could be exposed. During the impact, the vertical pressure distribution was observed using several pressure cells installed at the sand–slab interface. A total of 35 tests were carried out, systematically combining sand layer thickness ($D$), block’s equivalent diameter ($B$), and free-fall drop height ($H$), related to impact velocity. The masses of the released blocks were in the range of 117 to 7,399 kg, corresponding to diameters in the range of 0.42 to 1.79 m. Five free-fall drop heights up to 33 m were considered to reach impact velocities up to $90\mathrm{km}/\mathrm{h}$, covering the range of most velocities observed in actual rockfall studies. Three thicknesses of the sand layer protecting the thick SRC slab were considered: 1, 1.5, or 2 m. Data reduction from this full-scale impact tests program makes it possible to characterize, for a given thickness of protective sand layer, the time-space pressure pulse distribution applied to the protected structure during the impact for a large range of rock boulder masses and speeds actually observed in the field.

Methane and other gases released from soft sediments are among the main sources of greenhouse gases in the atmosphere. In this paper, a growth model for isolated bubbles in the sediments was established based on the theory of linear elastic fracture mechanics. Water level drop experiments were conducted using magnesium lithium philip silicate transparent soils, and the changes in bubble pressure and morphology during water level drop were analyzed. The experimental results show that there is a critical pressure for bubble growth caused by a drop in water level. Bubbles only start to grow by fracturing the overlying sediments when the water level drops to the critical value because the critical bubble pressure is lower than the actual bubble pressure. The strength of soil, depth of the bubble position, longitudinal length of bubble, and amount and rate of water level drop are key factors affecting isolated bubble growth. Bubbles in the soils with higher strength are more difficult to reach the critical state but have a faster growth rate once they do reach it. The depth of bubble position only affects the time reaching the critical state and does not impact the post-growth process. Deeper bubbles are more difficult for initiating growth. For bubbles at the same depth, larger bubbles begin growing earlier. As bubbles become larger, the growth rate of the bubble increases progressively faster. Faster water level drops result in shorter times to reach their critical state and accelerate their growth rate.

Pile installation leads to significant changes in soil state (i.e., void ratio and effective stress) around the pile, which affects stiffness and bearing capacity. Currently, the driveability of piles is analyzed using empirical methods, and the ultimate bearing capacity is estimated without considering the installation effects. This paper presents simulations of the entire installation and subsequent axial bearing capacity of a close-ended pile using a single numerical tool based on the material point method (MPM). A lab-scale experiment is used as a validation case, where the pile is first impact-driven in dry sand, with different initial relative densities (from loose to very dense), and then axially loaded. A state-dependent constitutive model (DeltaSand) is used in the numerical simulations to predict the mechanical behavior of the sand at different relative densities with a single set of input parameters. The paper also illustrates several enhancements needed to obtain more accurate results: (1) an improved contact algorithm that allows gap closure; (2) a rigid-body formulation for the pile body; and (3) a general analytical solution for calculation of energy-consistent impact forces in uncoupled hammer-pile systems.

The filter paper (FP) method, a cost-effective, versatile, and convenient approach, has found extensive application in analyzing fine-grained soils, enabling the successful measurement of a broad spectrum of total and matric suctions. However, using the FP method in sand with a limited range of suction profiles has been notably underrepresented in the existing literature. Therefore, this study investigated the applicability of the contact FP method in establishing soil water retention curves (SWRCs) using five distinct FP calibration curves for 10 poorly graded sands. Among these calibration curves, the one presented by Chandler yielded the highest suction values, which were followed by those of ASTM. Furthermore, the FP results were subjected to fitting with a mathmatical model, which led to the observation that the overall trends in the obtained SWRCs displayed smooth sigmoidal shapes, resulting in a complete and coherent SWRCs. The SWRC results of five sand samples were assessed through validation against the hanging column technique (HCT) results, indicating a high degree of concordance and affirming the suitability of the FP method for sands characterized by a limited suction profile. Notably, the entire range of suction values fell within the 0.01–20 kPa range based on the calibration curves presented by Chandler and ASTM for all the examined sand samples, except sands 9 and 10. Furthermore, the suction values were unobtainable using the two aforementioned calibration curves within the ranges of 0.01–1.47 kPa and 0.01–0.172 kPa.

Spectral analysis of surface waves (SASW) and multichannel analysis of surface waves (MASW) are the two main active surface wave methods, and their common inversion schemes are effective mode (EM) inversion and fundamental mode (FM) inversion, respectively. The former has a more rigorous inversion scheme, which considers receiver locations and dynamic response, while the latter has a more convenient dispersion analysis and fast inversion process. This paper introduces an improved dynamic response solution for elastic layered media subjected to vertical load. It is fast, accurate, and unconditionally stable, facilitating the full-wavefield inversion in terms of the frequency–velocity spectrum (FVS). The MASW FVS inversion merges the convenience and robustness of MASW dispersion analysis with the rigorous inversion scheme adopted by SASW. It is based on the actual dynamic response at receiver locations and takes into account possible higher modes and nonRayleigh waves. In this study, various types of synthetic velocity profiles are used to compare the inversion results among three different inverted schemes (SASW EM, MASW FM, and MASW FVS inversion). The results indicate that, when phase-unwrapping in SASW is not problematic, SASW EM inversion and MASW FVS inversion are theoretically sounder and produce better inverted results than the MASW FM inversion, while the process of MASW FVS inversion is more convenient and robust. Finally, a field example is used to further demonstrate the advantages of FVS inversion and its applicability.